Foundations TCP/IP

(1)

Foundations TCP/IP

4370.book Page i Friday, July 23, 2004 1:02 PM

(2)

4370.book Page ii Friday, July 23, 2004 1:02 PM

(3)

San Francisco ◆ London

Foundations TCP/IP

Andrew G. Blank

4370.book Page iii Friday, July 23, 2004 1:02 PM

(4)

Associate Publisher: Neil Edde Acquisitions Editor: Heather O’Connor Developmental Editor: Heather O’Connor Production Editor: Rachel Gunn

Copyeditor: Anamary Ehlen

Compositor: Craig Woods, Happenstance Type-O-Rama Graphic Illustrator: Tony Jonick, Rappid Rabbit Proofreaders: Laurie O'Connell, Nancy Riddiough Indexer: Lynnzee Elze

Book Designer: Judy Fung Cover Design: Ingalls + Associates Cover Photo: Jerry Driendl, Taxi

Copyright © 2004 SYBEX Inc., 1151 Marina Village Parkway, Alameda, CA 94501. World rights reserved. No part of this publication may be stored in a retrieval system, transmitted, or reproduced in any way, including but not limited to photo- copy, photograph, magnetic, or other record, without the prior agreement and written permission of the publisher.

Library of Congress Card Number: 2004109311 ISBN: 0-7821-4370-9

SYBEX and the SYBEX logo are either registered trademarks or trademarks of SYBEX Inc. in the United States and/or other countries.

Screen reproductions produced with Collage Complete and FullShot 99. FullShot 99 © 1991–1999 Inbit Incorporated. All rights reserved. FullShot is a trademark of Inbit Incorporated. Collage Complete is a trademark of Inner Media Inc.

TRADEMARKS: SYBEX has attempted throughout this book to distinguish proprietary trademarks from descriptive terms by following the capitalization style used by the manufacturer.

The author and publisher have made their best efforts to prepare this book, and the content is based upon final release software whenever possible. Portions of the manuscript may be based upon pre-release versions supplied by software manufacturer(s). The author and the publisher make no representation or warranties of any kind with regard to the completeness or accuracy of the contents herein and accept no liability of any kind including but not limited to performance, merchantability, fitness for any particular purpose, or any losses or damages of any kind caused or alleged to be caused directly or indirectly from this book.

Manufactured in the United States of America 10 9 8 7 6 5 4 3 2 1

4370.book Page iv Friday, July 23, 2004 1:02 PM

(5)

To my inspiration, my encourager, my perfect match, my best friend, and the love of my life, my wife, Suzie, you have had a profound and awesome impact on my life. I love you very much.

To my son, A.J. and my daughter, Amber, I treasure your love and have tremendous pride in both of you; Daddy loves you so much.

4370.book Page v Friday, July 23, 2004 1:02 PM

(6)

Acknowledgments

Several people have assisted me in many ways while writing this book. I’d like to acknowledge their contribu- tions and offer my sincere appreciation.

I appreciate several devoted people at Sybex. I have had the privilege of working closely with some very talented people, especially Rachel Gunn and Heather O’Connor. Anamary Ehlen did an exceptional job of editing my garbled-up thoughts into complete sentences. Many thanks to Sybex production department, including proofreaders Laurie O'Connell and Nancy Riddiough, indexer Lynnzee Elze, and compositor Craig Woods at Hap- penstance Type-O-Rama, who diligently turned text into print. I applaud the imagination and creativity of Tony Jonick in turning my sketches into illustrations. What an awesome honor to work with all of you!

I’d like to acknowledge the encouragement and prayers of my family and friends. All things are possible!

4370.book Page vi Friday, July 23, 2004 1:02 PM

(7)

Introduction

When you’re learning any new topic or technology, it’s important to have all of the basics at your disposal. The Sybex Foundations series provides the building blocks of specific technologies that help you establish yourself in IT.

TCP/IP is the de facto protocol of the Internet, and this protocol is supported by every major network operating system. As more organizations and individuals connect networks and computers to the Internet and one another, there is a continuing need for IT professionals to have a thorough understanding of this protocol suite. TCP/IP Foundations assumes no prior knowledge of TCP/IP and provides a solid introduction to this core networking topic, explaining the fundamentals of TCP/IP in simple terms with tangible examples.

My goal with TCP/IP Foundations is to introduce you to TCP/IP concepts so that you’ll come away with an intermediate understanding of TCP/IP. This book isn’t boringly technical; each topic is covered to sufficient depth, but not to an extreme.

As a network administrator and instructor, I have several years’ experience working in the computer industry and specifically with TCP/IP. Pulling from this experience, I’ve tried to present the relevant material in an interesting way, and I’ve included what I have found to be the most important concepts. The book is filled with several simple examples, diagrams, and screen captures in an effort to make the TCP/IP protocol more tangible.

This book is neither operating system–specific nor software-specific. Concepts are presented so that you can gain an understanding of the topic without being tied to a particular platform.

Who Should Read This Book?

TCP/IP Foundations is designed to teach the fundamentals of the TCP/IP protocol stack to people who are fairly new to the topic. This book will be useful for:

◆ People interested in learning more about TCP/IP

◆ Decision-makers who need to know the fundamentals in order to make valid, informed choices around TCP/IP

◆ Administrators who feel they are missing some of the foundational information about TCP/IP

◆ Small business owners interested in the protocol they will likely use on their networks

◆ Those interested in learning more about how data moves across the Internet

◆ Instructors teaching a TCP/IP fundamentals course

◆ Students enrolled in a TCP/IP fundamentals course

4370.book Page xiii Friday, July 23, 2004 1:02 PM

(14)

xiv Introduction

What This Book Covers

Working with TCP/IP has been an interesting, exciting, and rewarding experience. As I continue to learn about computers and TCP/IP, the more I see the need to continue learning. No matter what sector of the computer industry you’re employed in (or even if you’re not employed in IT yet), TCP/IP is an important foundational topic that you must understand; TCP/IP is the current and future standard protocol for networking.

TCP/IP Foundations contains many drawings and charts that help create a comfortable learning environment. It provides many real-world analogies that you will be able to relate to and through which the TCP/IP protocol will become tangible. The analogies provide a simple way to understand the technical process that is occurring through TCP/IP.

This book continues to build your understanding about TCP/IP progressively, like climbing a ladder. Here’s how the information is presented:

Chapter 1 This chapter provides an overview of where TCP/IP and the Internet came from and how they are related.

Chapters 2–5 These chapters describe what a protocol is and what the OSI and DoD models are. These chapters include a discussion of what happens at each layer in the DoD model and why the model is important.

Chapters 6–10 These chapters describe TCP/IP addressing—what IP addresses look like and how they are implemented. You’ll learn how to assign IP addresses both manually and through Dynamic Host Configura- tion Protocol (DHCP). You’ll learn all about DHCP. You’ll also learn about subnet masks—what they are, what they do, and how to create them.

Chapters 11–14 These chapters focus on name resolution methods and implementations. You’ll learn why name resolution is needed and the steps you need to take to resolve names. You’ll learn about Domain Name System (DNS), Dynamic DNS, and Windows Internet Naming Service (WINS).

Chapter 15 You’ll learn about the future of TCP/IP—the transition to a new version of IP in the next few years. This chapter gives you a heads-up on what to expect, and tells you how to find out more.

Making the Most of This Book

At the beginning of each chapter of TCP/IP Foundations, you’ll find a list of the topics I’ll cover within the chapter.

custom subnet mask

A nonstandard subnet mask used by a network administrator to make more efficient use of a network address by creating more subnets.

To help you soak up new material easily, I’ve highlighted new terms, such as custom subnet mask, in italics and defined them in the page margins.

And to give you some hands-on experience, you’ll find Test It Out sections in the chapters that allow you to practice what you’ve just learned. In addition, several special elements highlight important information:

4370.book Page xiv Friday, July 23, 2004 1:02 PM

(15)

Introduction xv

Notes provide extra information and references to related information.

Tips are insights that help you perform tasks more easily and effectively.

Warnings let you know about things you should—or shouldn’t—do as you learn more about TCP/IP.

At the end of each chapter, you can test your knowledge of the chapter’s relevant topics by answering the review questions. (You’ll find the answers to the review questions in Appendix A.)

There’s also some special material for your reference. If you’re wondering what certain acronyms stand for, Appendix B is an acronym guide spelling out the acronyms used in this book. If you’d like to quickly look up the meaning of a term, the Glossary has all the terms that have been introduced throughout the book.

4370.book Page xv Friday, July 23, 2004 1:02 PM

(16)

4370.book Page xvi Friday, July 23, 2004 1:02 PM

(17)

In This Chapter

Chapter 1 The Origin of TCP/IP and the Internet

Two people can communicate effectively when they agree to use a common language. They could speak English, Spanish, French, or even sign language, but they must use the same language.

Computers work the same way. Transmission Control Protocol/Internet Protocol (TCP/IP) is like a language that computers speak. More specifically, TCP/IP is a set of rules that defines how two computers address each other and send data to each other. This set of rules is called a protocol. Multiple protocols that are grouped together form a protocol suite and work together as a protocol stack.

TCP/IP is a strong, fast, scalable, and efficient suite of protocols. This protocol stack is the de facto protocol of the Internet. As information exchange via the Internet becomes more widespread, more individuals and companies will need to understand TCP/IP.

◆The features of TCP/IP

◆ARPAnet

◆TCP’s method of moving data

◆Requests For Comments (RFCs)

◆The benefits of using TCP/IP 4370c01.fm Page 1 Tuesday, July 13, 2004 6:03 AM

(18)

2 Chapter 1

What Is TCP/IP?

protocols

Rules or standards that govern communications.

TCP/IP is a set of protocols that enable communication between computers.

There was a time when it was not important for computers to communicate with each other. There was no need for a common protocol. But as computers became networked, the need arose for computers to agree on certain protocols.

network administrator

A person who installs, monitors, and troubleshoots a network.

Today, a network administrator can choose from many protocols, but the TCP/IP protocol is the most widely used. Part of the reason is that TCP/IP is the protocol of choice on the Internet—the world’s largest network. If you want a computer to communicate on the Internet, it’ll have to use TCP/IP.

When multiple protocols work together, the group is collectively known as a protocol suite or protocol stack. TCP/IP is an example of a protocol suite (it describes multiple protocols that work together). The implementation of TCP/IP is described as a protocol stack. Both terms are used interchangeably, yet their definitions vary slightly.

Another reason for TCP/IP’s popularity is that it is compatible with almost every computer in the world. The TCP/IP stack is supported by current versions of all the major operating systems and network operating systems—including Windows 95/98, Windows NT, Windows 2000, Windows XP, Windows 2003, Linux, Unix, and NetWare.

Unlike proprietary protocols developed by hardware and software vendors to make their equipment work, TCP/IP enjoys support from a variety of hardware and software vendors. Examples of companies that have products that work with TCP/IP include Microsoft, Novell, IBM, Apple, SuSE, and Red Hat. Many other companies also support the TCP/IP protocol suite.

TCP/IP is sometimes referred to as “the language of the Internet.” In addition to being the official language of the Internet, TCP/IP is also the official language of many smaller networks. For all the computers that are attached to the Internet to communicate effectively, they must agree on a language. Just as every human language has certain rules so that the people involved in the conversation understand what the other is saying, a computer language needs a set of rules so that computers can effectively communicate. Some of the rules of a language that computers use to communicate include determining when to send data and when to receive data.

Features of TCP/IP

TCP/IP has been in use for more than 20 years, and time has proven it to be a tested and stable protocol suite. TCP/IP has many features and benefits. In this section, you will learn about some of the most important ones.

4370c01.fm Page 2 Tuesday, July 13, 2004 6:03 AM

(19)

The Origin of TCP/IP and the Internet 3

Support from Vendors

As stated earlier, TCP/IP receives support from many hardware and software vendors. This means that the TCP/IP suite is not tied to the development efforts of a single company. Instead, the choice to use TCP/IP on a network can be based on the purpose of the network and not on the hardware or software that has been purchased.

Interoperability

host

Any device (such as a workstation, server, mainframe, or printer) on a network or internetwork that has a TCP/IP address.

One of the major reasons why the TCP/IP suite has gained popularity and accep- tance so universally is that it can be installed and used on virtually every platform. For example, using TCP/IP, a Unix host can communicate and transfer data to a DOS host or a Windows host. A host is another name for a computer or device on a network. TCP/IP eliminates the cross-platform boundaries.

Flexibility

TCP/IP is an extremely flexible protocol suite, and in later chapters you will learn about some features that contribute to this flexibility. Examples of TCP/IP’s flexibility include the latitude an administrator has in assigning and reassigning addresses. An administrator can automatically or manually assign an IP address to a host, and a TCP/IP host can convert easy-to-remember names, such as www.sybex.com, to a TCP/IP address.

Routability

A limitation of many protocols is their difficulty in moving data from one segment of the network to another. TCP/IP is exceptionally well adapted to the process of routing data from one segment of the network to another, or from a host on a network in one part of the world to a host on a network in another part of the world.

In the following sections, you will learn about how these features of TCP/IP grew out of the military’s need for a reliable, flexible networking standard.

The Origins of the Internet: ARPAnet

Understanding the roots of the Internet will give you insight into the development of TCP/IP and many of its rules and standards. If you know why TCP/IP was created and how it evolved, the TCP/IP protocol suite is easier to understand.

ARPAnet

The Advanced Research Projects Agency’s supernetwork—the predecessor of the Internet.

The predecessor of today’s Internet was ARPAnet, a supernetwork that was created by the Advanced Research Projects Agency (ARPA) and launched in 1969. This network was created in response to the potential threat of nuclear attack from the Soviet Union. One of ARPA’s primary goals was to design a fault-tolerant network that would enable U.S. military leaders to stay in contact in case of nuclear war. By the standards of the time, this fault-tolerant network seemed to be almost science fiction. ARPA set out on a mission to create a network with what seemed to be impossible requirements.

(20)

4 Chapter 1

In the late 1950s, the United States Department of Defense (DoD), under the guid- ance of one of America’s leading think tanks, the RAND corporation, formed the Advanced Research Projects Agency (ARPA).

Network Control Protocol (NCP) The protocol used before TCP/IP.

The protocol, or language of choice, used on the ARPAnet was called Net- work Control Protocol (NCP)—TCP/IP had not yet been developed. As the ARPAnet grew, however, a new protocol was needed because NCP simply didn’t fulfill all the needs of a larger network. The NCP protocol was similar to a human language that has only a few words. The language might enable a few people to communicate, but as you include more people who want to talk about many more subjects, you have to improve the language.

The ARPAnet project had some specific goals and requirements. To reach these goals and meet these requirements, some of the top computer minds worked in a collaborative effort with little financial or public glory. Many of the top computer minds that worked on the ARPAnet were affiliated with major universities. It was not the intention of the project leaders to create the worldwide network that exists today, but fantastic growth soon followed the ARPAnet’s humble beginnings.

ARPAnet’s Requirements

To fulfill the needs of the military, the new ARPAnet had to meet the following requirements:

No one point more critical than any other Because the network needed to be able to withstand a nuclear war, there could be no one critical part of the network and no single point of failure. If there were any critical parts of the network, enemies could target that area and eliminate communications.

Redundant routes to any destination Because any location on the network could be taken down by enemies in the event of a war, there had to be multiple routes from any source to any destination on the network.

Without redundant routes, any one location could become a critical communications link and a potential point of failure.

On-the-fly rerouting of data If any part of the network failed, the network had to be able to reroute data to its destination on-the-fly.

Ability to connect different types of computers over different types of networks This network could not be tied to just one operating system or hardware type. Because universities, government agencies, and cor- porations often rely on different types of Local Area Networks (LANs) and network operating systems, interoperability among these many networks was critical. Connecting to the network should not dictate that a lot of new hardware had to be purchased; rather, the existing hardware should suffice.

(21)

Not controlled by a single corporation If one corporation had a monop- oly on this network, the network would grow to boost the corporation instead of the usefulness and effectiveness of the network. This network needed to be a cooperative effort among many engineers who were working to improve the network for the sake of the supernetwork, not that of a corporation.

By December of 1969 the ARPAnet had four hosts. The ARPAnet consisted of computers at the University of California at Los Angeles, the University of California at Santa Barbara, the University of Utah, and Stanford Research Insti- tute. The ARPAnet set the foundation for what would grow up to be the Internet.

Requests For Comments

Request For Comments (RFC) A paper thoroughly describing a new protocol or technology.

To improve the technology that was being used on the ARPAnet, a system was designed to encourage and facilitate correspondence among the engineers who were developing this new network. This system, which is still in use today, relies on Requests For Comments (RFCs) to provide feedback and collaboration among engineers. An RFC is a paper that has been written by an engineer, a team of engineers, or just someone with a better idea, to define a new technology or enhance an existing technology.

The process of submitting RFCs was designed to be a “bulletin board” for posting technical theories. The old-school way of writing a thesis or book was too slow. RFCs provided an informal and fast way to share new technologies and ideas for enhancements. After an RFC is written and posted, it can be evaluated, critiqued, and used by other engineers and developers. If another engineer or developer can improve on the theory or standard, the RFC provides an open forum in which to do so. Many of these papers are long, painstakingly technical, and in most cases good reading material for someone with difficulty sleeping.

Internet Engineering Task Force (IETF) A governing body of the Internet.

An RFC can be submitted for review to the Internet Engineering Task Force (IETF). Engineers from the IETF review the papers that are submitted and assign a number to each. From that point on, the RFC number becomes the effective

“name” of the paper. For example, the first RFC, which is about host software, is called RFC 1. RFC 1 was submitted in 1969 by a developer named Steve Crocker. There are currently more than 3,000 RFCs.

As the ARPAnet was growing and researchers and engineers were making improvements, they used RFCs as a tool to strengthen and ensure the network’s foundation. TCP/IP is a child of the RFC method of development—no corporation makes money when you install TCP/IP. Using RFCs has been the method of growing the ARPAnet with the best network minds contributing.

It is possible for anyone to write and publish an RFC. Instructions on how to write and submit an RFC are detailed in RFC 2223. Today, RFCs are posted on many Web sites.

(22)

6 Chapter 1

The Birth of TCP/IP

As stated earlier, the “language” spoken by hosts on the ARPAnet in 1969 was called NCP. However, NCP had too many limitations and was not robust enough for the supernetwork, which was beginning to grow out of control. The limitations of NCP and the growth of the ARPAnet led to research and development of a new network language.

Transmission Control Protocol (TCP) The protocol describing communication between hosts.

In 1974 Vint Cerf and Bob Kahn, two Internet pioneers, published “A Proto- col for Packet Network Interconnection.” This paper describes the Transmission Control Protocol (TCP), which is a protocol in the protocol suite that would eventually replace NCP.

The TCP protocol describes the host-to-host portion of a communication.

TCP explains how two hosts can set up this communication and how they can stay in touch with each other as data is being transferred. NCP did not resolve these issues to the extent that TCP was able to.

As you will learn in later chapters, TCP is responsible for making sure that the data gets through to the other host. It keeps track of what is sent and retransmits anything that did not get through. If any message is too large for one package, TCP splits the message into several packages and makes sure that they all arrive correctly. After they have arrived, TCP at the other end puts all the packages back together in the proper order.

Transmission Control Protocol/

Internet Protocol (TCP/IP) The suite of protocols that when combined create the “language of the Internet.”

By 1978, testing and further development of this language led to a new suite of protocols called Transmission Control Protocol/Internet Protocol (TCP/IP). In 1982, it was decided that TCP/IP would replace NCP as the standard language of the ARPAnet. RFC 801 describes how and why the transition from NCP to TCP was to take place. On January 1, 1983, ARPAnet switched over to TCP/IP, and the network continued to grow exponentially.

In 1990, the ARPAnet ceased to exist. The Internet has since grown from ARPAnet’s roots, and TCP/IP has evolved to meet the changing requirements of the Internet.

TCP/IP Founda- tions

NCP

Language Course 4370c01.fm Page 6 Tuesday, July 13, 2004 6:03 AM

(23)

Design Goals of TCP/IP

TCP/IP has evolved to its current state. The protocols within the TCP/IP suite have been tested, modified, and improved over time. The original TCP/IP protocol suite had several design goals that intended to make it a viable protocol for the large, evolving internetwork. Some of these goals included:

Hardware independence A protocol suite that could be used on a Mac, PC, mainframe, or any other computer.

Software independence A protocol suite that could be used by different software vendors and applications. This would enable a host on one site to communicate with a host on another site, without having the same software configuration.

Failure recovery and the ability to handle high error rates A protocol suite that featured automatic recovery from any dropped or lost data. This protocol must be able to recover from an outage of any host on any part of the network and at any point in a data transfer.

Efficient protocol with low overhead A protocol suite that had a minimal amount of “extra” data moving with the data being transferred. This extra data, called overhead, functions as packaging for the data being transferred and enables the data transmission. Overhead is similar to an envelope used to send a letter, or a box used to send a bigger item—having too much overhead is as efficient as using a large crate to send someone a necklace.

Ability to add new networks to the internetwork without service disruption A protocol suite that enabled new, independent networks to join this network of networks without bringing down the larger internetwork.

Routable Data A protocol suite on which data could make its way through an internetwork of computers to any possible destination. For this to be possible, a single and meaningful addressing scheme must be used so that every computer that is moving the data can compute the best path for every piece of data as it moves through the network.

The TCP/IP protocol suite has evolved to meet these goals. Throughout this book, you will learn how TCP/IP has met and surpassed these original design goals.

Moving Data across the Network

Creating this new “supernetwork” introduced many new concepts and challenges for the pioneering engineers. One of the most critical issues was how to move data across the network. Older communications protocols relied on a circuit-switched technology. TCP/IP, however, introduced a new way of moving data across a network. The protocol suite set a new standard for communications and data transport by using a packet-switched network.

(24)

8 Chapter 1

TCP/IP’s method of moving data and information helped the protocol suite fulfill several of the requirements for the growing ARPAnet supernetwork. In the following sections, you’ll learn about how circuit-switched and packet-switched communications methods work.

Moving Data on a Circuit-Switched Network

circuit-switched network

A network on which all data in a communication takes the same path.

Historically, data has moved through a circuit-switched network. In a circuit- switched network, data moves across the same path throughout the entire communication. An example of a circuit-switched network is the telephone system.

When you make a telephone call, a single path (also called a circuit) is established between the caller and the recipient. For the rest of the conversation, the voice data keeps moving through the same circuit. If you were to make a call and get a very staticky connection, you would hang up and try again. This way you could get a different circuit, hopefully one with less static. Early network data transmissions followed this type of pathway.

In the illustration below, notice that although the data could take multiple routes, all the data moves from the source to the destination along the same path.

In a circuit-switched network, data communication moves along a single, established route.

Moving Data on a Packet-Switched Network

A circuit-switched network was unacceptable for both the ARPAnet and the Internet. Data had to be able to move through different routes so that if one circuit went down or got staticky, it didn’t affect communication on the rest of the network. Instead, data simply would take a different route.

R=Router D=Data

R

R R

R

D D D

D

Source Destination

(25)

packet-switched network

A network on which the data in a communication takes several paths.

The Internet uses a packet-switched network. On a packet-switched network, the computer that is sending the data fragments the data into smaller, more manageable chunks. These chunks are called packets. Each packet is then individu- ally addressed and sent to its intended recipient. As the several packets make their way through the network, each packet finds its own way to the receiver.

The receiving computer reassembles the packets into the original message.

packet

A unit of data that is prepared for transmission onto a network.

The illustration below shows how TCP/IP moves data. Notice that there are several routes that the data packets can follow from the source to the destination.

Unlike the illustration on the preceding page, the data packets here use a variety of routes—some follow the same path, while others follow different paths. Each packet follows its own route, and data is reassembled at the destination. This is how information moves on a packet-switched network.

Understanding How a Packet-Switched Network Functions

To help you understand how a packet-switched network moves data, let’s look at a similar real-world situation.

Let’s say that I take my son’s soccer team to an arcade and restaurant for a team party. I have the whole team outside of the arcade. My task is to get the team to the other side of the arcade, to my wife who is waiting for them in the restaurant. In this analogy, the team represents the complete file on one host, and each child represents a data packet. One of my goals is to lose as few of the kids as possible.

While we are standing outside, it is easy to put the team in order; all the children are wearing numbered jerseys. I tell the kids that we will meet on the other side of the arcade in a restaurant for pizza and that they should all move as fast as possible

Continues through the arcade and to the restaurant.

R=Router D=Data

R

R R

R

D D

D

D D

Destination Source

(26)

10 Chapter 1

Why Use TCP/IP?

TCP/IP offers many advantages over other network protocols and protocol suites.

Here is a summary of some of the benefits of using the TCP/IP protocol suite:

Widely published, open standard TCP/IP is not a secret. It is not proprietary or owned by any corporation. Because it is a published protocol with no secrets, any computer engineer is able to improve or enhance the protocol by publishing an RFC.

Compatible with different computer systems TCP/IP enables any system to communicate with any other system. It is like a universal language that would enable people from any country to communicate effectively with people from any other country.

Works on different hardware and network configurations TCP/IP is accepted and can be configured for virtually every network created.

Routable protocol TCP/IP can figure out the path of every piece of data as it moves through the network. Because TCP/IP is a routable protocol, the size of any TCP/IP network is virtually unlimited.

Reliable, efficient data delivery TCP/IP can guarantee that the data is transferred to another host.

After I open the door and say, “Go,” the kids enter one at a time. Entering the arcade one at a time represents the fragmenting and sending of the file. Just as each of the kids has a numbered jersey, each packet has a number so that the receiving host can put the data back together.

Now picture a dozen six-year-olds moving through the arcade. Some of the children will take a short route; others will take a long route. Possibly, they’ll all take the same route, though it is much more likely that they will all take different routes. Some will get hung up at certain spots, but others will move through faster. My wife is in the restaurant waiting to receive the team. As they start arriving at the restaurant, she can reassemble the children (packets) in the correct order because they all have a number on their backs. If any are missing, she will wait just a bit for the stragglers and then send back a message that she is missing part of the team (file).

After I receive a message that she is missing a child (a packet), I can resend the missing part. I do not need to resend the entire team (all the packets), just the missing child (packet or packets).

Please note, however, I would not go look for the lost child; I would just put the same numbered jersey on a clone of the lost child and send him into the arcade to find the restaurant.

(27)

Single addressing scheme TCP/IP uses a single and relatively simple addressing scheme. You will learn about TCP/IP’s addressing in Chapter 6,

“IP Addressing.” An administrator can transfer knowledge of TCP/IP to any TCP/IP network without relearning the addressing scheme.

The Internet has become a necessity for business, and it soon will be a necessity at home. Many businesses, large and small, are connected to the Internet and are using TCP/IP as the protocol of choice for their internal networks. As more and more homes connect to the Internet, those computers will also use the TCP/IP protocol suite. The commercial implications of the Internet have changed the dynamic of every business model that has ever been taught.

internetwork

Several smaller networks connected together.

TCP/IP is the standard for a communications protocol on the Internet.

You cannot connect to the Internet without using TCP/IP. Whether you build a network at home with two hosts or you manage an internetwork at your business with 100,000 hosts, TCP/IP is a communications protocol that will work effectively. TCP/IP can scale to any size environment and is robust enough to connect different types of LANs.

These are a few of the many reasons why network administrators choose to use TCP/IP as the protocol on their networks.

Terms to Know

ARPAnet packets

circuit-switched network packet-switched network

host protocols

Internet Engineering Task Force (IETF)

Requests For Comments (RFCs)

internetwork Transmission Control

Protocol (TCP)

network administrator Transmission Control Protocol/

Internet Protocol (TCP/IP) Network Control Protocol (NCP)

(28)

12 Chapter 1

Review Questions

1. The Internet was originally called:

2. List three requirements that the military mandated of this new network.

3. Another name for a computer on a TCP/IP network is:

4. Describe packet-switched and circuit-switched networks.

5. What is an RFC?

6. What protocol did TCP/IP replace?

7. True or False: TCP/IP is one protocol.

8. What is IETF?

9. List four benefits of using TCP/IP.

10. What year was the change made from NCP to TCP/IP?

(29)

In This Chapter

Chapter 2 Protocols

In the first chapter, you learned how the Internet grew from the ARPAnet and how TCP/IP was developed. As the computer network industry has grown, rules and standards have evolved. These rules and standards have formed the TCP/IP protocol into a popular and robust standard used by computers to communicate. This chapter examines why protocols are important and how they enable communication between hosts.

◆Protocols

◆Packets

◆The seven layers of the OSI model

◆The four layers of the DoD model 4370c02.fm Page 13 Tuesday, July 13, 2004 6:04 AM

(30)

14 Chapter 2

What Are Protocols?

A protocol is a rule or a set of rules and standards for communicating that computers use when they send data back and forth. Both the sender and receiver involved in data transfer must recognize and observe the same protocols.

To exchange data, the sending and the receiving computers, also called hosts, must agree on what the data will look like. When one host is sending another host a whole bunch of 1s and 0s, both hosts have to agree on the meaning and placement of each 1 and each 0. Part of the information that is sent represents addresses and part is data—each host has a unique address, just as you have a unique address on your street. And just like a letter being delivered to your address, data is delivered to the appropriate host based on its address. The hosts that send the information must understand how to find the correct address among the data so that the data can be routed to its destination.

When hosts begin communicating with each other, they first must agree on what protocols to use. This is similar to two people who are going to have a conversation: They have to agree on which language to use and what the rules for the conversation will be. They must agree on who will talk first, how to address the other, how to acknowledge that the information is understood, and how to finish or close the conversation. In the following illustration, Harry the Host is trying to set up communication with another host. The first thing that they need to agree on is the language, or protocols, to use.

protocol suite

A combination of protocols.

A group of protocols is called a protocol suite or a protocol stack. A single protocol addresses one particular issue that helps to enable communication—for example, defining what an address looks like. When combined with other protocols, the protocol group that results is called a protocol suite. TCP/IP, for example, is a protocol suite. At a computer that is communicating on a network, the software that packages the data and prepares it for transmission is called a protocol stack. When a computer is receiving data, the data moves up through the protocol stack.

Harry Gary

Dutch English

Italian English

Spanish French 4370c02.fm Page 14 Tuesday, July 13, 2004 6:04 AM

(31)

Protocols 15

protocol stack

Protocols that send and receive data.

Protocol suites are typically referred to by just a couple of the protocols in the suite. Rather than refer to a suite by a name that might include as many as 20 protocols, you can simply reference it by an easier-to-use and more friendly name. Many protocol suites are in use today. Some are proprietary protocols that have limited use. These are developed for specific purposes to meet some particular need of the hardware or software involved.

Some of the popular protocol suites in today’s network communications include:

IPX/SPX This is the protocol suite that Novell has implemented with its operating system. The acronym stands for Internetwork Packet Exchange/

Sequenced Packet Exchange.

AppleTalk This is the protocol suite that Apple has implemented with its operating system.

TCP/IP This is the protocol suite that has been made a standard of the Inter- net. Anyone who would like to use the Internet must use the TCP/IP suite.

Some of the questions that a protocol might answer include:

◆ What type of cable or transmission media is used to connect hosts on the network?

◆ How is data transmitted on the transmission media?

◆ How do the hosts on the network know when to transmit data?

◆ How does each host know how much data can be transmitted at a time?

◆ How can hosts using different operating systems communicate?

◆ How can a host check the data received for transmissions?

Protocols Move Packets of Data

When data is sent from one host to another, the Transmission Control Protocol of TCP/IP divides the data into more manageable “chunks.” As explained in Chapter 1, “The Origin of TC/IP and the Internet,” these chunks are called packets. The protocol determines how the packets are formed and addressed—the packets are like crates that are used to ship the data.

headers

Bits of information attached to each packet that usually include addressing and routing details; the information acts like a little sticky note on the packet.

Each of the packets has a set of headers applied to it. The headers usually include addressing and routing information, which makes it possible to reassemble the packets and have the original data at the destination. The headers are applied to the packets for the same reason that you’d apply labels to a package that you are sending. Several headers may be applied to each packet.

A host sending data to another host is like me sending a package to some- body else—for instance, sending a bicycle to my sister in another state. The bicycle represents data that is going to be transferred to another host. To send the bicycle, I have to follow certain rules, or protocols. I put the bicycle into a package, or maybe more than one package if it doesn’t fit into a single package.

In this example, the packages represent packets.

(32)

16 Chapter 2

Even after the bicycle is inside the packages, it is not going anywhere until I put some addressing information on it. There are protocols for putting addresses on the packages: I must use my sister’s correct name as well as her correct address. The address label must include the pieces of information necessary to get the packages to the correct destination—for example, her street address, city, state, and zip code. This is similar to TCP/IP putting addressing information headers on the packets that are being transmitted. I also put my return address on the labels, which is similar to a data packet including its source information.

There is a proper place for all this addressing information, and I must correctly fill it in on every package or it will not get there. Finally, I indicate the order in which to open the packages by writing “1 of 6,” “2 of 6,” etc. on them. This will let my sister know which package to open first, second, and so on so that she can easily reassemble the bike.

encapsulation

The wrapping of a packet into the appropriate package or format.

After the packages are ready to go, I need to decide which delivery service to use. The packages’ format depends on the delivery service I choose: If I use Federal Express, I will put the packages into FedEx boxes; if I use United Parcel Service, I will put the packages into a UPS format. Similarly, packets are encapsulated into a format that is appropriate for the physical network that the sending host is located on. If the host is on an Ethernet network, the packet must be in the appropriate format to travel on an Ethernet network. If it’s on a Token Ring network, it must be in the Token Ring format. Encapsulation is a fancy word for wrapping up the packet into the appropriate package or format.

Because I’m on a UPS route, I call Mike, the UPS man, and ask him to pick up the packages. Neither Mike nor I actually deliver the packages. Instead, the data, packaged in the appropriate format, moves through the transport system, being transferred from one location to the next. The packages might take different routes, but they will get to the same destination. They are delivered to the destination based on the address that I put on the labels. If there is a problem with the delivery, the system will let me know because I put my return address on the packages.

After the packages arrive, my sister opens them. She can reassemble the bicycle based on the information that was on the labels. Similarly, the recipient of the data packet can assemble the data based on the information in the packets’ headers.

My sister discards the packing material after she uses the pertinent information from the labels. All she really wants is the bicycle; the packaging was used only to send the bicycle to the correct destination and in the correct order. When using TCP/IP to transport data, a packet is built with several headers, which are discarded after the important information has been used and the data has been delivered to the requesting application.

The illustration on the next page shows Harry the Host sending data to Sally the Host. Notice that the data has been fragmented into several packets and that

(33)

Protocols 17

each packet includes sequence numbers. As the receiving host, Sally reassembles the data back to its unfragmented format.

Why We Need Protocols and Standards

Rules—or protocols and standards—are important to ensure compatibility between different kinds of things. As more and more hardware and software vendors began joining the technology explosion, there was no guarantee that any of their products would be able to work with one another. A system had to be put in place so that hardware and software consumers would not get burned by buy- ing incompatible systems.

For example, let’s say that I own a small business and I want to buy some new computer equipment. I go out and find some hardware and software that will make my business run smoother and more effectively. All the vendors tell me how great their hardware and software is, so I buy it. I’ve been sold the dream of how my new automated office will function and how I’ll have nothing but spare time. I’ve been told that everything works together and that my small business will be successful as a result.

Sally From: Harry

To: Sally 4 of 4

EED

1 of 4 THE D

2 of 4 ATA Y Harry From: Harry

To: Sally 3 of 4 OU N 4370c02.fm Page 17 Tuesday, July 13, 2004 6:04 AM

(34)

18 Chapter 2

However, I bought some hardware from one vendor, some software from another, some other hardware from another vendor, and more software from yet another. And guess what? None of the stuff works together. I just spent a ton of money, and now I’m spending all my time calling for support. All the nice support people are telling me it’s the other vendor’s software or hardware that is causing the problem.

To keep this scenario from happening, standards and protocols were developed. If the hardware and software vendors were all working with the same guidelines—the same standards and protocols—then their hardware and software should all work together. The hardware vendor would continue to make money selling his hardware, the software vendor would continue to make money selling his software, and I would make money in my small- to medium-sized automated business. I would be happy to buy more hardware and software because it works and it serves my purposes.

Developing protocols is an ongoing, ever-changing science. New protocols are constantly under development and testing, and they are improved as the need arises. As the industry is increasing so dynamically and rapidly, more protocols are unleashed to handle the boom. However, before a protocol is accepted and widely implemented, it has to pass rigorous testing. A standard framework is used to help design, compare, test, and evaluate protocols.

The OSI Reference Model

International Organization for Standardization (ISO) The organization that ratified the OSI model.

For network communications to take place, hundreds of questions must be answered by a set of protocols. Evaluating and working with these hundreds of questions would be unmanageable. So, in 1977 the International Organization for Standardization (ISO) adopted the Open Standards Interconnection (OSI) model. The OSI model breaks down the many tasks involved in moving data from one host to another. Now instead of having hundreds of questions to answer, the OSI model gives us a reference to work with. The hundreds of questions are divided into seven smaller, more manageable groups of questions. The seven groups are called layers.

Open Standards Interconnection model (OSI)

A seven-layer model used to break down the many tasks involved in moving data from one host to another.

The OSI reference model is exactly that; it is only a model. If we continue to think of the model as a set of questions that have to be answered, then the protocols are the answers. Any one protocol may answer only a few of the questions or, in other words, address specific layers in the model. By combining multiple protocols into a protocol suite, we can answer all the questions posed by the model.

layer

A portion of the OSI model that is used to categorize specific concerns.

The OSI model was created by first making a list of most computer networking topics, such as routing, reliability, and sequencing. From this list, all of the topics were categorized by how they are used in network communications. Within each layer, several topics are discussed. Breaking down this huge task of data communication into seven layers makes the task more manageable.

(35)

Protocols 19

The seven layers of the OSI model are explained in the following sections.

The OSI reference model functions as a baseline for comparison to any protocol suite. As such you can use the OSI model—or the DoD model, which you’ll learn about later in this chapter—to help you understand how the parts of TCP/IP work.

This baseline function of the OSI model is similar to a model home. When designing your new home, a model can be used as a baseline. Everyone in the neighborhood also uses the model home as reference to help make the choices in the new homes that they are building. All the homes will vary slightly from the model, but the model provides a means for comparison. In the same way, you can compare any protocol suite to the OSI reference model because protocols are designed from this model. The OSI model acts as a baseline for creating and com- paring networking protocols.

The Seven Layers of the OSI Model

The goal of the OSI model is to break down the task of data communication into simple steps. These steps are called layers, and the OSI model is made up of seven distinct layers. Each layer has certain responsibilities.

The seven layers of the OSI model are:

◆ Application

◆ Presentation

◆ Session

◆ Transport

◆ Network

◆ Data-Link

◆ Physical

You will learn about the responsibilities of each of these layers in the following sections. The OSI model is a method of compartmentalizing data-communication topics in a way that can help a network administrator when troubleshooting.

Responsibilities of Each Layer

The purpose of each layer in the OSI model is to provide services to the layer above it while shielding the upper level from what happens below. The higher layers do not need to know how the data got there or what happened at the lower layers.

The following illustration shows how data moves through the seven layers of the OSI model. Here, Harry the Host is transmitting data onto a network. He

(36)

20 Chapter 2

could be saving a file from his word processing application to a file server, for example. As the data moves down the seven layers toward the network, each layer puts a little bit of information called a header on the packet. The exact contents of each header depend on the protocols enabled at each layer of the protocol suite.

What’s Your Favorite Layer of the OSI Model?

Here’s an interesting party topic and excellent conversation starter. Recently I had a heated discussion with a colleague that lasted almost an hour. We were arguing about which is our favorite layer of the OSI model, and I was amazed at how fast we dug in our heels to defend which layer and why. I found myself deeply loyal to the Physical layer, while my colleague had the opinion that the Presentation layer is best. My point was that all of the important “blue-collar” stuff happens at the Phys- ical layer. The Physical layer works down in the trenches getting bits onto the wire and taking them off. He pointed out that the Presentation layer is so important because it uses compression and encryption. As the discussion got more heated, I found myself thinking of the Presentation layer as a wimpy layer while building up the many important tasks that the Physical layer handles!

Since this discussion, I teach that this is actually a tremendous way to learn the OSI model. Find another network administrator and defend your favorite layer. Come up with valid reasons why you like and don’t like each layer. Then take turns defending different layers.

Application Presentation

Session Transport

Network Data-Link

Data

Data Data Data Data Data

Data

Physical

Data 4370c02.fm Page 20 Tuesday, July 13, 2004 6:04 AM

(37)

Protocols 21

The Application Layer

The top layer of the OSI model is the Application layer. The purpose of the Application layer is to manage communications between applications. A standard Application layer program such as FTP or SMTP interacts with a program that is running at the local workstation. The programmer who has written a word processing application writes the program to interact with a standard application that exists at the Application layer. The word processor uses the standard network application to save, copy, or delete files. This is the layer where the applications receive data and request data. All other layers work for this layer. Think of the Application layer as the CEO of the OSI model.

The Presentation Layer

The Presentation layer is the layer below the Application layer and above the Session layer. The Presentation layer adds structure to packets of data being exchanged. The primary job of the Presentation layer is to ensure that the message gets transmitted in a language or syntax that the receiving computer can understand. The protocols at the Presentation layer may translate the data into an understandable syntax and then compress and maybe encrypt the data before passing it down to the Session layer. Some people may choose this as their favorite layer because it presents the data to the Application layer, and the Application layer is so important.

The Session Layer

The Session layer is below the Presentation layer. It controls the dialog during communications. The Session layer protocols set up sessions, or connections.

These protocols cover such topics as how to establish a connection, how to use a connection, and how to break down the connection when a session is com- pleted. After a connection is established, the Session layer protocols check for transmission errors. The Session layer also adds control headers to the data packets during the exchange of data.

The Transport Layer

Below the Session layer is the Transport layer. The Transport layer can guarantee that packets are received. The Transport layer also can establish a connection and send acknowledgments as packets are received. The protocols in this layer provide the means to establish, maintain, and release connections for the hosts involved in communication.

The Network Layer

The Network layer, which is below the Transport layer, is responsible for routing the packet based on its logical address. The Network layer fragments and reassembles packets if necessary. It also moves the packets of data from the source to the

Foundations TCP/IP

Foundations TCP/IP

Foundations TCP/IP

Andrew G. Blank

Acknowledgments

Contents

Introduction

Who Should Read This Book?

What This Book Covers

Making the Most of This Book

Chapter 1

The Origin of TCP/IP and the Internet

What Is TCP/IP?

Features of TCP/IP

Support from Vendors

Interoperability

Flexibility

Routability

The Origins of the Internet: ARPAnet

ARPAnet’s Requirements

Requests For Comments

The Birth of TCP/IP

Design Goals of TCP/IP

Moving Data across the Network

Moving Data on a Circuit-Switched Network

Moving Data on a Packet-Switched Network

Why Use TCP/IP?

Terms to Know

Review Questions

Chapter 2

Protocols

What Are Protocols?

Protocols Move Packets of Data

Why We Need Protocols and Standards

The OSI Reference Model

The Seven Layers of the OSI Model

Responsibilities of Each Layer

Application Presentation

Session Transport

Network Data-Link

Physical

The Application Layer

The Presentation Layer

The Session Layer

The Transport Layer

The Network Layer